PURPOSE: We aimed to assess the performance of computed tomography (CT) in localizing site of traumatic gastrointestinal tract (GIT) injury and determine the diagnostic value of CT signs in site localization. METHODS: CT scans of 97 patients with surgically proven GIT or mesenteric injuries were retrospectively reviewed by radiologists blinded to surgical findings. Diagnosis of either GIT or mesenteric injuries was made. In patients with GIT injuries, site of injury and presence of CT signs such as focal bowel wall hyperenhancement, hypoenhancement, wall discontinuity, wall thickening, extramural air, intramural air, perivisceral infiltration, and active vascular contrast leak were evaluated. RESULTS: Out of 97 patients, 90 had GIT injuries (70 single site injuries and 20 multiple site injuries) and seven had isolated mesenteric injury. The overall concordance between CT and operative findings for exact site localization was 67.8% (61/90), partial concordance rate was 11.1% (10/90), and discordance rate was 21.1% (19/90). For single site localization, concordance rate was 77.1% (54/70), discordance rate was 21.4% (15/70), and partial concordance rate was 1.4% (1/70). In multiple site injury, concordance rate for all sites of injury was 35% (7/20), partial concordance rate was 45% (9/20), and discordance rate was 20% (4/20). For upper GIT injuries, wall discontinuity was the most accurate sign for localization. For small bowel injury, intramural air and hyperenhancement were the most specific signs for site localization, while for large bowel injury, wall discontinuity and hypoenhancement were the most specific signs. CONCLUSION: CT performs better in diagnosing small bowel injury compared with large bowel injury. CT can well predict the presence of multiple site injury but has limited performance in exact localization of all injury sites.
PURPOSE: We aimed to assess the performance of computed tomography (CT) in localizing site of traumatic gastrointestinal tract (GIT) injury and determine the diagnostic value of CT signs in site localization. METHODS: CT scans of 97 patients with surgically proven GIT or mesenteric injuries were retrospectively reviewed by radiologists blinded to surgical findings. Diagnosis of either GIT or mesenteric injuries was made. In patients with GIT injuries, site of injury and presence of CT signs such as focal bowel wall hyperenhancement, hypoenhancement, wall discontinuity, wall thickening, extramural air, intramural air, perivisceral infiltration, and active vascular contrast leak were evaluated. RESULTS: Out of 97 patients, 90 had GIT injuries (70 single site injuries and 20 multiple site injuries) and seven had isolated mesenteric injury. The overall concordance between CT and operative findings for exact site localization was 67.8% (61/90), partial concordance rate was 11.1% (10/90), and discordance rate was 21.1% (19/90). For single site localization, concordance rate was 77.1% (54/70), discordance rate was 21.4% (15/70), and partial concordance rate was 1.4% (1/70). In multiple site injury, concordance rate for all sites of injury was 35% (7/20), partial concordance rate was 45% (9/20), and discordance rate was 20% (4/20). For upper GIT injuries, wall discontinuity was the most accurate sign for localization. For small bowel injury, intramural air and hyperenhancement were the most specific signs for site localization, while for large bowel injury, wall discontinuity and hypoenhancement were the most specific signs. CONCLUSION: CT performs better in diagnosing small bowel injury compared with large bowel injury. CT can well predict the presence of multiple site injury but has limited performance in exact localization of all injury sites.
Authors: A Furukawa; M Sakoda; M Yamasaki; N Kono; T Tanaka; N Nitta; S Kanasaki; K Imoto; M Takahashi; K Murata; T Sakamoto; T Tani Journal: Abdom Imaging Date: 2005 Sep-Oct
Authors: V Maniatis; H Chryssikopoulos; A Roussakis; C Kalamara; S Kavadias; A Papadopoulos; J Andreou; K Stringaris Journal: Abdom Imaging Date: 2000 Jul-Aug
Authors: Stefan Wirth; Julian Hebebrand; Raffaella Basilico; Ferco H Berger; Ana Blanco; Cem Calli; Maureen Dumba; Ulrich Linsenmaier; Fabian Mück; Konraad H Nieboer; Mariano Scaglione; Marc-André Weber; Elizabeth Dick Journal: Insights Imaging Date: 2020-12-10